Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Looking for an APL implementation???

52 views
Skip to first unread message

jski

unread,
Dec 16, 2011, 1:53:41 PM12/16/11
to
I'm looking for an APL implementation for either or both Windows or
Linux or Cygwin?

I work with a group of mechanical engineers in Huntsville, AL on
vibration problems - so this is definitely intended for high end
numerical problems.

Is A+ the ticket?

What about Dyalog?

Some other commercial product?

Some academic exercise?

---John

Aaron W. Hsu

unread,
Dec 16, 2011, 3:52:54 PM12/16/11
to
On Fri, 16 Dec 2011 13:53:41 -0500, jski <john.chl...@gmail.com>
wrote:

> I'm looking for an APL implementation for either or both Windows or
> Linux or Cygwin?

The best development environment on Linux is APLX I believe. I use Dyalog
right now for other reasons, but I have a copy of both Dyalog and APLX and
I can recommend them both. Both have 64-bit solutions as well, making them
good for large computations.

--
Aaron W. Hsu | arc...@sacrideo.us | http://www.sacrideo.us
לֵ֤ב חֲכָמִים֙ בְּבֵ֣ית אֵ֔בֶל וְלֵ֥ב כְּסִילִ֖ים בְּבֵ֥ית שִׂמְחָֽה

jski

unread,
Dec 16, 2011, 5:08:07 PM12/16/11
to
On Dec 16, 3:52 pm, "Aaron W. Hsu" <arcf...@sacrideo.us> wrote:
> On Fri, 16 Dec 2011 13:53:41 -0500, jski <john.chludzin...@gmail.com>
> wrote:
>
> > I'm looking for an APL implementation for either or both Windows or
> > Linux or Cygwin?
>
> The best development environment on Linux is APLX I believe. I use Dyalog
> right now for other reasons, but I have a copy of both Dyalog and APLX and
> I can recommend them both. Both have 64-bit solutions as well, making them
> good for large computations.
>
> --
> Aaron W. Hsu | arcf...@sacrideo.us |http://www.sacrideo.us
> לֵ֤ב חֲכָמִים֙ בְּבֵ֣ית אֵ֔בֶל וְלֵ֥ב כְּסִילִ֖ים בְּבֵ֥ית שִׂמְחָֽה

What's the best Windows solution?

jski

unread,
Dec 16, 2011, 5:07:01 PM12/16/11
to
On Dec 16, 3:52 pm, "Aaron W. Hsu" <arcf...@sacrideo.us> wrote:
> On Fri, 16 Dec 2011 13:53:41 -0500, jski <john.chludzin...@gmail.com>
> wrote:
>
> > I'm looking for an APL implementation for either or both Windows or
> > Linux or Cygwin?
>
> The best development environment on Linux is APLX I believe. I use Dyalog
> right now for other reasons, but I have a copy of both Dyalog and APLX and
> I can recommend them both. Both have 64-bit solutions as well, making them
> good for large computations.
>
> --
> Aaron W. Hsu | arcf...@sacrideo.us |http://www.sacrideo.us
> לֵ֤ב חֲכָמִים֙ בְּבֵ֣ית אֵ֔בֶל וְלֵ֥ב כְּסִילִ֖ים בְּבֵ֥ית שִׂמְחָֽה

How about IBM's APL2? From what I gather, it's available on both Win7
and Linux?

---John

Aaron W. Hsu

unread,
Dec 16, 2011, 9:31:10 PM12/16/11
to
On Fri, 16 Dec 2011 17:08:07 -0500, jski <john.chl...@gmail.com>
wrote:

> What's the best Windows solution?

I don't think there is one best solution.

--
Aaron W. Hsu | arc...@sacrideo.us | http://www.sacrideo.us

Aaron W. Hsu

unread,
Dec 16, 2011, 9:31:37 PM12/16/11
to
On Fri, 16 Dec 2011 17:07:01 -0500, jski <john.chl...@gmail.com>
wrote:

> How about IBM's APL2? From what I gather, it's available on both Win7
> and Linux?

I imagine that APL2 is also a good system, but I have never tried it, so I
cannot comment on it.

--
Aaron W. Hsu | arc...@sacrideo.us | http://www.sacrideo.us

jski

unread,
Dec 17, 2011, 10:06:06 AM12/17/11
to
On Dec 16, 9:31 pm, "Aaron W. Hsu" <arcf...@sacrideo.us> wrote:
> On Fri, 16 Dec 2011 17:08:07 -0500, jski <john.chludzin...@gmail.com>

> wrote:
>
> > What's the best Windows solution?
>
> I don't think there is one best solution.
>
> --
> Aaron W. Hsu | arcf...@sacrideo.us |http://www.sacrideo.us

> לֵ֤ב חֲכָמִים֙ בְּבֵ֣ית אֵ֔בֶל וְלֵ֥ב כְּסִילִ֖ים בְּבֵ֥ית שִׂמְחָֽה

OK.

How about a couple of good Windows solutions?

---John

Chris Bruce

unread,
Dec 17, 2011, 1:20:06 PM12/17/11
to
I have experience only of APL2 on mainframe and PC and so cannot
usefully comment on the other "dialects" on the PC platform. However...

...you can freely download the APL2 manuals from
http://www-01.ibm.com/software/awdtools/apl/library.html?S_CMP=rnav to
enable you to see what features are offered by IBM's implementation of
APL, such as external routines, auxiliary and attached processors, and
the supplied workspaces including graph plotting. I believe you can
download a trial version of APL2 from the site too via
http://www-01.ibm.com/software/awdtools/apl/.

You may like to take a look at the British APL Association's website for
some good stuff: http://sites.google.com/site/baavector/

There is also the APL wiki which also has some good stuff:
http://aplwiki.com/

Good luck with your search for a solution.

jski

unread,
Dec 17, 2011, 3:32:13 PM12/17/11
to
On Dec 17, 1:20 pm, Chris Bruce <chris@nospam_mirrorcomputing.co.uk>
wrote:
> On 16/12/2011 22:07, jski wrote:
>
>
>
> > On Dec 16, 3:52 pm, "Aaron W. Hsu"<arcf...@sacrideo.us>  wrote:
> >> On Fri, 16 Dec 2011 13:53:41 -0500, jski<john.chludzin...@gmail.com>
> >> wrote:
>
> >>> I'm looking for an APL implementation for either or both Windows or
> >>> Linux or Cygwin?
>
> >> The best development environment on Linux is APLX I believe. I use Dyalog
> >> right now for other reasons, but I have a copy of both Dyalog and APLX and
> >> I can recommend them both. Both have 64-bit solutions as well, making them
> >> good for large computations.
>
> >> --
> >> Aaron W. Hsu | arcf...@sacrideo.us |http://www.sacrideo.us
> >> לֵ֤ב חֲכָמִים֙ בְּבֵ֣ית אֵ֔בֶל וְלֵ֥ב כְּסִילִ֖ים בְּבֵ֥ית שִׂמְחָֽה
>
> > How about IBM's APL2?  From what I gather, it's available on both Win7
> > and Linux?
>
> > ---John
>
> I have experience only of APL2 on mainframe and PC and so cannot
> usefully comment on the other "dialects" on the PC platform. However...
>
> ...you can freely download the APL2 manuals fromhttp://www-01.ibm.com/software/awdtools/apl/library.html?S_CMP=rnavto
> enable you to see what features are offered by IBM's implementation of
> APL, such as external routines, auxiliary and attached processors, and
> the supplied workspaces including graph plotting. I believe you can
> download a trial version of APL2 from the site too viahttp://www-01.ibm.com/software/awdtools/apl/.
>
> You may like to take a look at the British APL Association's website for
> some good stuff:http://sites.google.com/site/baavector/
>
> There is also the APL wiki which also has some good stuff:http://aplwiki.com/
>
> Good luck with your search for a solution.

I've written to IBM asking about ALP2 on Windows or Linux but haven't
heard back. By chance, do you know if it's available on Windows or
Linix? Their web site seems to indicate this but that doesn't
necessarily mean anything.

---John

Aaron W. Hsu

unread,
Dec 17, 2011, 4:06:40 PM12/17/11
to
On Sat, 17 Dec 2011 10:06:06 -0500, jski <john.chl...@gmail.com>
wrote:

> How about a couple of good Windows solutions?

The three I would be looking at is Dyalog, APLX, and APL2.

--
Aaron W. Hsu | arc...@sacrideo.us | http://www.sacrideo.us

kai

unread,
Dec 17, 2011, 4:47:44 PM12/17/11
to
For commercial applications Dyalog, APLX and APL2000 are the ones to
look at.

APL2 might still be fines if one needs to crunch numbers and nothing
else, or has to communicate with a mainfraime, but APL2 has not seen
any major new feature for many years now.

For non-commercial stuff NARS2000 is certainly fine.

Björn Helgason

unread,
Dec 17, 2011, 5:12:47 PM12/17/11
to

Chris Bruce

unread,
Dec 18, 2011, 4:11:19 AM12/18/11
to
Yes - it works on Windows; that's what I should have said instead of "PC"!

You could try posting your question(s) on IBM's APL2 forum
(https://www.ibm.com/developerworks/forums/forum.jspa?forumID=2104).
I've always found the support from the APL2 development team very good
so you may get an answer more quickly from them.

rusi

unread,
Dec 18, 2011, 10:50:09 AM12/18/11
to
On Dec 16, 11:53 pm, jski <john.chludzin...@gmail.com> wrote:
> I'm looking for an APL implementation for either or both Windows or
> Linux or Cygwin?
>
> I work with a group of mechanical engineers in Huntsville, AL on
> vibration problems - so this is definitely intended for high end
> numerical problems.
>
> Is A+ the ticket?

If you want to try aplus under linux you may want to look at
http://www.emacswiki.org/emacs/AplInDebian

[Id be interested to know if you tried and it did not work]

Kerry Liles

unread,
Dec 19, 2011, 10:16:19 AM12/19/11
to
APL2 is no doubt a fine product (you can download it as a trial version
- it will run for a reasonable length of time for evaluation purposes...
up to a max threshold of cpu time if I recall) BUT, the price for the
average person is quite prohibitive I think.

Ric

unread,
Dec 19, 2011, 9:55:32 PM12/19/11
to
Depending on what exactly you mean by APL, I'd recommend J.
Available (at no cost) on Windows, Linux and Mac. Certainly well
suited for high-end numerical problems.
http://www.jsoftware.com
http://www.jsoftware.com/jwiki/Essays

Don Kelly

unread,
Dec 20, 2011, 1:41:18 AM12/20/11
to
"jski" wrote in message
news:fe9efe09-90a1-483d...@h4g2000yqk.googlegroups.com...
--------------------------------dk responds----------------------------
In effect, you are looking at something that may be routine in electrical
power engineering- a combination of circuit analysis and matrix methods
setting up state space equations. Now you consider non -linear behaviour.
You may have a set of non-linear differential equations to deal with-you
will need an iterative approach- as in the case of power system load flows.

Define your problem and look at some of the electrical analogies that are
available.
Concentrate on the methods to deal with the problem. Yes, it may involve
solving a set of state equations and using the result to solve another set
(linear or otherwise set of simultaneous equations, and iterating. This is
very similar to consideration of power system stability analysis. Both parts
can be handled well by APL but possibly even better by J.
{example a load flow program can be written in APL as less than 20-25
lines - of which the core calculations take 5 -8 lines- the rest is
comment or iteration control. J can shave this a bit and can inherently
make use of sparse matrices).



Don Kelly
cross out to reply


0 new messages